Post #677
clem 🤗 @ClementDelangue
Introducing Qwen3.8-27B-pi-adapted 🤗 We adapted Qwen's model specifically for the pi coding agent using reinforcement learning and a LoRA adapter. The results, measured as pass@1: 🚀 SWE-bench Pro (Hard 51): 29.4% → 62.7% 💻 Terminal-Bench 2.1: 39.5% → 55.8% ⚡ 55–62% fewer tool calls per task This is the power of open models: you can adapt them to YOUR tools and YOUR use cases. Let’s build! 💛 Open weights: huggingface.co/harness-hub/Qw… Methodology: huggingface.co/blog/harness-h…
Open on X ↗· 0 views · 0 clicks
Join the conversation with the author on X.
SHARE YOUR PICK
Put it on your timeline.
Post the link on X. This card shows up as the preview.
